Method for detecting a movement of a body

The method enhances eye movement detection by filtering and thresholding image pairs to extract relevant images, addressing inefficiencies in existing technologies by reducing the computational load and improving detection precision for rapid eye movements.

EP4402648B1Active Publication Date: 2024-12-04P3LAB
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
EP2023790701
Authority / Receiving Office
EP · EP
Patent Type
Patents
Current Assignee / Owner
Priority Date
2022-10-24
Filing Date
2023-10-23
Publication Date
2024-12-04
Estimated Expiration
2043-10-23

AI Technical Summary

Technical Problem

Current methods for detecting eye movements, such as saccades, are inefficient due to the need to process large numbers of images, with approximately 80% not showing pixel variations, leading to high computational demands and inefficiencies.

Method used

A method that determines a similarity measure between image pairs, applies a filtered derivative to eliminate static components, and extracts images exceeding a threshold, allowing for efficient detection of body movements by processing only relevant images.

Benefits of technology

This approach significantly reduces the number of images that need to be processed, improving detection efficiency and enabling faster and more precise tracking of eye movements, particularly for rapid movements like saccades.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure IMGF0001
    Figure IMGF0001
  • Figure IMGF0002
    Figure IMGF0002
  • Figure IMGB0001
    Figure IMGB0001
Patent Text Reader

Abstract

The invention relates to a computer-implemented method for detecting a movement of a body from a sequence of images of the body by isolating the images (51) belonging to pairs of images in which a function (3) derived from a measurement (2) of the similarity of the pairs of images is greater than a predetermined threshold (4). Thus the body movement is detected on these images (51), which can then be processed more precisely for the purpose of studying the movement, without the need to process all the images in the image sequence.
Need to check novelty before this filing date? Find Prior Art

Description

Technical field

[0001] The present invention relates to: a method for detecting a movement of a body, an eye tracking method, a use of the method for detecting an ocular saccade, a computer data processing system, a device for detecting a movement of a patient's eye, a computer program, and a computer-readable computer medium. Prior art

[0002] For neurological assessments of a patient, it is known to subject the patient's eyes to visual stimuli and monitor their response to these stimuli. These stimuli are, for example, presented in the form of one or more pixels on a screen. One or more cameras capture a sequence of images of the patient's eyes simultaneously with the presentation of these stimuli. This sequence of images is then processed to track the patient's eye movement.

[0003] Saccades define a known class of movements that can be detected. A "saccade" is a rapid redirection of the line of sight from one center of interest to another. It is a very rapid eye movement that can reach a speed of 600 degrees per second. In order to effectively detect and track a saccade, it is therefore necessary for the camera acquisition rate to be sufficiently high, for example, around 800 frames per second.

[0004] The number of images to be processed during this neurological assessment is therefore quite considerable. Their processing requires high-power computing tools as well as a certain amount of execution time.

[0005] In the separate context of eye blink detection, the article "Eye Blink Detection Using Local Binary Patterns" by K. Malik and B. Smolka, ICMCS, IEEE, from 2014, discloses a method for transforming eye images into local binary patterns between which a distance is determined and then regularized. The resulting signal includes peaks that are known to correspond to the desired eye blinks. While this method is suitable for the case of an eye blink, it does not allow for effective detection and tracking of a saccade. Statement of the invention

[0006] An object of the invention is to make such processing more efficient. For this purpose, the invention provides a method for detecting a movement of a body, for example an eye, from a sequence of images of this body, the method being implemented by computer and comprising the following steps: (i) determining a similarity measure of each pair of images in a sequence of pairs of images, the sequence of pairs of images being induced by the sequence of images and including each image of the latter; (ii) determining a derivative function of the similarity measure on the sequence of pairs of images, the derivative function corresponding to a filtered derivative obtained by combination with a low-pass filter; (iii) extracting at least one image from each pair of images in which the derivative function is greater than a strictly positive threshold; (iv) detecting body movement based on the images extracted in step (iii).

[0007] This process allows for faster and more efficient detection of body movements in the image sequence, and therefore, ultimately, more efficient processing of these images, particularly when there are a large number of them.

[0008] Indeed, rather than processing all the images in the sequence for the purpose of detecting body movement, the inventor proposes to extract the images in the sequence initially containing information about body movement, and to detect this movement on these images. These images can further be processed, for example by segmentation, to follow or analyze this movement in more detail. Thus, not all the images need to be considered in this regard, which significantly improves the efficiency of body movement detection, and more generally of image processing.

[0009] This method is particularly motivated, in the context of the prior art, by the fact that approximately 80% of the images captured by the camera do not show any variation in pixels compared to the preceding image in the sequence of images and therefore do not correspond to eye movements. To improve the tracking of eye movement, it was therefore desirable not to process these images but only the images from which eye movement could be detected, which is what the method according to the invention allows. Of course, this is not limited to this application. Those skilled in the art will understand that it can be used for any detection of body movement in a sequence of images. The lower the proportion of images on which the movement to be detected, the more advantageous the method according to the invention is.

[0010] To determine from which images to detect body movement, and therefore which images to extract in step (iii), the method includes steps (i) and (ii) which are at the heart of the invention.

[0011] First, in step (i), a similarity is measured between images. The images are ordered in pairs for this purpose. This measurement makes it possible to determine a degree of similarity between the two images of each pair, typically pixel by pixel. However, this step is not sufficient on its own to determine the images to be extracted in step (iii). Indeed, two images may present more or less significant variations in pixels without the body represented on the images having undergone any movement. These variations are, for example, due to noise and / or variable brightness of the body's environment during image acquisition. Step (ii) aims to derive the similarity measure used in step (i) to eliminate the static component of the images in terms of pixels and better distinguish a real movement of the body from the aforementioned variations. Thus, the extraction of step (iii) is more refined and better meets the efficiency needs sought in this context.

[0012] The threshold used in step (iii) advantageously allows the proportion of extracted images to be adapted. It corresponds to a compression level of the image sequence. A zero threshold is obviously not desirable because it would correspond to retaining all the images regardless of their representation. The threshold is set according to the type of similarity measure and derivative function used.

[0013] The detection of the body movement can then be done in step (iv) by processing the extracted images. In one embodiment, the detection takes the form of a simple proof of the existence of a body movement which emerges at first sight from the images by step (iii). Processing of the extracted images can be carried out during step (iv) to detect and / or characterize more precisely the body movement as known to those skilled in the art, the essential thing remaining that this step is carried out on a reduced number of images.

[0014] By using the terms "computer-implemented", the method of the invention involves the use of a computer, a computer network and / or any other programmable (e.g., a smartphone, a tablet, an FPGA, etc.) or programmed (e.g., an integrated circuit / ASIC, etc.) device. In particular, the term "computer" cannot be interpreted restrictively. The determination steps therefore fall at least partially within an underlying computer nature. For example, one or more of these steps may constitute a determination by algorithmic calculation.

[0015] For the purposes of this document, the term "body" is used generically and can refer to either a physical object or all or part of a human body.

[0016] For the purposes of this document, the images are preferably of the same size, for example 640 x 480 pixels. The method of the invention can however be applied selectively to a subset of the pixels of the image as detailed below. For the purposes of this document, the images are preferably captured at the same frequency, for example, at 797 images per second.

[0017] The term "sequence" itself implies a notion of order between the elements of the sequence. Preferably, the order associated with the sequence of images is that in which the images were captured, for example by a video camera. For example, in this case, the sequence of images forms a sequence of images in time, representing, for example, a movement of a part of a human body, for example of an eye. Other notions of order may be used.

[0018] The image pairs from step (i) are also ordered since they are classified in sequence. This sequence is "induced" by the image sequence, that is, the notion of order on the pairs is induced by that on the images. In other words, a first pair {I 1 , I 2} precedes a second pair {I 1 ', I 2 '} in the sequence of pairs, with I 1 and I 1 ' preceding respectively I 2 and I 2 ' in the image sequence, if and only if, in the image sequence: I 1 precedes or corresponds to I 1 ' and I 2 precedes or corresponds to I 2 '. The images in the same pair are obviously preferably distinct. All images in the image sequence are found in at least one of the image pairs. The similarity between the images in the image sequence can thus be measured successively according to the order induced by the image sequence.

[0019] Two preferred examples of image pair sequences are provided: according to a first embodiment, the sequence of pairs of images comprises (and preferably consists of) each pair of two consecutive images in the sequence of images; according to a second embodiment, the sequence of pairs of images comprises (and preferably consists of) each pair formed of the first image of the sequence of images and another image of the sequence of images.

[0020] In other words, if the image sequence corresponds to I n 1 ≤ n ≤ N = I 1 , I 2 , I 3 , … , I N where N denotes the number of images said first and second realizations correspond respectively to the sequences of pairs I n , I n + 1 1 ≤ n ≤ N − 1 et I 1 , I n 2 ≤ n ≤ N In the first embodiment, the similarity between images is measured progressively by two consecutive images over the entire image sequence (image 1 with image 2, then image 2 with image 3, and so on). Thus, it is possible to detect very precisely a lack of similarity between two images that follow each other in the image sequence. In the second embodiment, the similarity is measured relative to the first image, so that it is possible to determine from which image a body movement could be detected.

[0021] The use in this document of the verb "to understand", its variants or conjugations, does not exclude the presence of elements other than those mentioned. Similarly, the use of the indefinite article "un", "une", or the definite article "le", "la" or "l'", to introduce an element does not exclude the presence of a plurality of these elements.

[0022] The use of "similarity measures" to quantify a difference between the pixels of two images is known to those skilled in the art. In particular, there are several such measures that can be used in the context of the invention, for example by a correlation measure. However, in the context of the invention, an advantageous preferred similarity measure of pairs of images is identified. It corresponds to a Bhattacharyya coefficient (denoted "BC") of two discrete probability distributions associated with the images of the pair of images and defined from pixels of these images.

[0023] Specifically, if p and q denote the probability distributions defined on a set P of the pixels of the two images of the pair of images, the Bhattacharyya coefficient is defined by: BC p q = ∑ x ϵ P p x q x . The set of pixels may correspond to all the pixels in the image or to a subset thereof, such as selecting only one row and / or one column out of two. Although it is generally the natural logarithm of the inverse of this coefficient which is used as a measure of similarity, in the context of the invention, it is the coefficient itself which is used, in particular for the purposes of carrying out step (ii).

[0024] There are also several ways known to those skilled in the art of associating a probability distribution with the pixels of an image to evaluate this coefficient. In one embodiment of the invention, the discrete probability distribution p associated with each image is defined at each pixel x of the image by: p x = G x ∑ y ϵ P G y where G denotes a gray level function defined on the pixels of the image. In other words, p(x) is defined as a quotient of a gray level at pixel p by a sum of gray levels at each pixel of the image. The gray level at a pixel is typically evaluated numerically and discretized, for example on an integer scale from 0 to 255.

[0025] Advantageously, the use of the Bhattacharyya coefficient compared to other similarity measures is independent of the brightness of the images, which makes the method according to this embodiment more robust from step (i), in particular with a view to specifically detecting body movements.

[0026] Step (ii) may also be implemented in different ways using several discrete derivatives known to those skilled in the art. In particular, the term "derivative function" is used equivalently to "derivative" in this document, as would be understood by a person skilled in the art. In view of the discrete domain for which the derivative is determined, it is a filtered derivative. In particular, a low-pass filter is composed with the derivation of the similarity measure in order to eliminate or reduce possible noise on the images. The derivative function therefore corresponds to a filtered derivative obtained by combination with a low-pass filter.

[0027] The filtered derivative preferably corresponds to a discretization, preferably bilinear, of a Laplace transform of transfer function given by H(s) = s (s τ +1) -1< where τis a constant (more precisely, time constant of the low-pass filter combined with the derivation). This discretization allows the use of this regular expression in a digital system. The fact that the discretization is bilinear is preferred and advantageous in order to guarantee adequate stability of the low-pass filter between its regular form given by the transfer function and its discretized form.

[0028] Concretely, if S n corresponds to the similarity measure of the n-th pair of images of the sequence of pairs of images, if S' n corresponds to the evaluation of the function derived from the similarity measure on this n-th pair of images, and if t denotes a constant duration between two images captured by a camera, the derivative filtered according to this embodiment is defined recursively by S ′ n = 2 S n − S n − 1 2 τ + t + 2 τ − t 2 τ + t S ′ n − 1 . The initial step S' 1 = S 1 is typically used. The constant τis a real number, preferably an integer multiple of the sampling period t, this multiple being, for example, between 2 and 10, for example 5.

[0029] Returning to the general statement of the method according to the invention, it can be noted that the application (or evaluation) of the function derived from the similarity measure on the sequence of pairs of images naturally provides a data distribution. Preferably, the threshold used in step (iii) is defined by a percentile of this distribution. This is an advantageously practical way of compressing the sequence of images adequately by extracting a specific percentage of the images according to the object of the motion detection. Preferably, this percentile is chosen between the eightieth and the ninety-ninth percentile, which makes the image processing significantly more efficient since less than one image in five is extracted from the image sequence.In the case where the method aims at the detection of ocular saccades, it was found by the inventor that the use of the ninety-fifth percentile provided entirely satisfactory results because saccades are very brief and rapid movements. In the particular case of the filtered derivative of the Bhattacharyya coefficient based on a gray level in the pixels of the images, for . τ = 5t and t = 3.75 ms, it was shown that the corresponding threshold of 0.00175 advantageously made it possible to efficiently detect all the ocular saccades of any patient, by extracting at most 5% of the images from the image sequence, whatever the execution conditions of the method.

[0030] According to a preferred embodiment of the invention, the image that is extracted from a pair in step (iii) comprises the image of the highest rank in the image sequence among the two images of the pair of images. This makes it possible to retain for step (iv) and / or the possible subsequent processing the images in which "something is happening", in other words, the images where a start of movement is likely to be detected. In the case of a pair of the form {I n , I n+1}, the image I n+1 is therefore extracted, and in the case of a pair of the form {I 1 , I n}, the image I n is extracted, if the evaluation of the function derived from the similarity measure in this pair is greater than the threshold.

[0031] Both images of the pair can optionally be extracted.

[0032] The detection of body movement can be done in step (iv) solely on the basis of the extracted images. However, other images can be used, for example, for reasons of security or reliability of the method. Thus, according to a preferred embodiment of the invention, step (iii) also comprises a periodic extraction of an image from the sequence of images. This extraction can be done at a frequency of between 20 and 40 Hz, for example 30 Hz, for a usual image acquisition speed by a camera. Alternatively, one image every 20 to 40 images, for example every 30 images, can be extracted. Since the proportion of additional images thus extracted is low, taking these images into account does not significantly affect the efficiency of the method while advantageously strengthening its reliability and robustness.

[0033] Preferably, the method also comprises a step of segmenting the images extracted in step (iii). This segmentation step may occur during step (iv) and / or simultaneously or subsequently thereto. It may be part of a subsequent “processing” of the extracted images, for example, in order to classify, study and / or quantify the detected body movements. The segmentation as such is a method known to those skilled in the art. It may be of any known type. It may, for example, be based on regions or contours at the level of the images of the sequence, and / or also on a classification of the pixels by intensity (for example, luminous, gray level, etc.). The segmentation may be done locally, on a defined area of ​​the images, or globally.

[0034] Preferably, in the case where the segmentation of each image extracted in step (iii) is defined by an image segmentation function, this segmentation step is implemented via the following sub-steps: (s1) determining an estimate of the segmentation function of each image extracted in step (iii); (s2) recursively determining the segmentation function of each image extracted in step (iii) from its estimate.

[0035] Thanks to this implementation, it is possible to segment the images extracted in step (iii) both quickly and efficiently. Indeed, the segmentation functions are determined recursively, and therefore deduced from each other. The order associated with the recursion is typically that induced by the image sequence. This avoids individual processing of the images, which is therefore longer and less efficient. This advantage is increased because each segmentation function is determined from its estimation (which can itself be determined recursively or not depending on the case), thus constituting an intermediate step that is simpler to execute than the direct segmentation of the images. Fewer calculations are thus necessary because the determination of a segmentation function is similar to the adjustment of a residual with respect to its estimation.The recursive aspect also increases the reliability of the determination of image segmentation functions because it is then not based solely on the estimates but also on the functions already determined.

[0036] The term "segmentation function" used above refers to a function defined on the pixels of the images associating a model and / or a structure of the pixels of the images, typically the pixels having a "non-zero" (or non-constant) image by the function. When such a model or such a structure can be highlighted at the pixel level of a first of the images extracted in step (iii) (algorithmically or manually), the segmentation step makes it possible to advantageously and recursively integrate this model or this structure on the other images extracted in step (iii) to follow and / or analyze its evolution and / or detect it efficiently, without processing each image independently.

[0037] The notations (s1) and (s2) should not be interpreted as substep (s1) being executed in full before substep (s2). These may be nested, as in the following preferred embodiment.

[0038] Preferably, the estimation of a segmentation function of a second image is determined in sub-step (s1) by a composition of a vector field with a segmentation function of a first image, the first image preceding the second image in the sequence of images, the segmentation function of the first image being determined beforehand in sub-step (s2), and the corresponding vector field in a displacement of pixels between the second and the first images. Preferably, the first image directly precedes the second image in the sequence of images extracted in step (iii). Preferably, each segmentation function estimation of an n-th image is determined in a similar manner from the segmentation function of the (n-1)-th image of the sequence of images extracted in step (iii).Furthermore, the terms "first" and "second" should not be interpreted above as being limited to the position of these images according to the order induced on the images extracted in step (iii) by the order corresponding to the image sequence.

[0039] The vector field corresponds to a form of comparison between two images and encodes a displacement that the pixels of the second image would have to undergo to return to a position associated with the first image. Preferably, it is calculated by optical flow.

[0040] The estimates of the segmentation functions, as a first-order approximation of these functions, are thus easily calculable, which further improves the efficiency and speed of execution of the segmentation step, and therefore of the method comprising it according to the invention. Furthermore, in the case of images of an eye for the purposes of eye tracking, this embodiment makes it possible to easily deduce an accurate estimate of the position of the eye or of a part of the eye to be tracked on each image extracted in step (iii) and thus to strongly limit an area of ​​interest of the image on which to identify the eye or the part of the eye via step (s2). The area is all the more limited, and therefore the estimation reliable, as the eye or the part of the eye to be tracked varies in a limited manner in position between two consecutive images extracted in step (iii), which is generally the case when the initial image sequence comprises a large number of images captured per second.In particular, this achievement is perfectly suited to the detection and study of ocular saccades.

[0041] The invention also provides an advantageous eye tracking method comprising the following steps: (a) providing a sequence of images of a patient's eye; (b) detecting eye movement from the sequence of images; where step (b) is implemented by means of the method according to the invention. In this case, the method is applied in the case where the "body" corresponds to the eye from the sequence of images provided in step (a). As is apparent from the statement of the method, eye tracking is provided based on the detection of step (b).

[0042] The method according to the invention allows more efficient eye tracking since the movement of the eye is detected in step (b) more efficiently and quickly. More generally, the embodiments and advantages of the method according to the invention are transposed mutatis mutandis to this eye tracking method.

[0043] Preferably, step (b) is implemented by means of the method when it comprises a step of segmenting the images extracted in step (iii). The above-mentioned segmentation step is then preferentially implemented at least in the vicinity of a representation of the iris of the eye on each of the images extracted in step (iii), and the method further comprises the following steps: (c) determining a position (or position information) of a limbus of the eye based on the segmentation of the images extracted in step (iii); (d) determining a position (or position information) of the eye based on the position (or position information) of the limbus of the eye determined in step (c).

[0044] The method according to this embodiment makes it possible to obtain rapid and precise eye tracking compared to the known methods of the prior art. Indeed, the segmentation of the images extracted in step (iii) is carried out recursively so that each segmentation is deduced from a segmentation estimate and from at least one of the previous segmentations, which allows for precise and rapid execution of the segmentation step of step (b). The segmentation of an image of the eye in the vicinity of the pixels associated with the iris of the eye makes it possible to highlight an elliptical shape in the case of contour segmentation, essentially corresponding to the limbus of the eye, the position of which can thus be advantageously deduced in step (c) according to the parameters defining the ellipse.

[0045] Furthermore, the method according to this embodiment proposes to first take into account the limbus to determine the position of the eye in step (d), and thus differs from most known methods which are distinctively based on data relating to the pupil to determine the position of the eye. Indeed, the determination of the position of the eye by these methods is less precise than the embodiment of the method according to the present invention. This is due to the fact that the determination of pupil position characteristics is affected by several errors due, for example, to the fact that the speed and acceleration of eye movements (which can reach respectively 600° / s and 35000° / s 2< during saccades) induces a relative movement between the pupil and the limbus of the eye, or to the fact that any black spot (for example, mascara) is taken into account in the calculations by most known eye trackers.

[0046] The position of the eye can however be further determined in step (d) on the basis of information on the position of the pupil of the eye previously determined in step (c) on the basis of the segmentation of the images extracted in step (iii), which makes it possible to obtain even more precise eye tracking.

[0047] The speed and efficiency of the eye tracking method also make it particularly interesting for tracking an eye over a continuous stream of video images. The initial image sequence is then provided in the form of this image stream, for example, by means of a camera pointed at the eye. The subject to which the eye belongs is typically stimulated to follow a moving target on a screen. The eye tracking method then allows for the study of eye movements, for example, for the purpose of detecting neurodegenerative diseases.

[0048] In particular, and in view of the advantages of the invention set out above, the invention specifically proposes a use of the eye tracking method according to the invention, for detecting a saccade of the eye, typically from a stream of video images of the eye constituting the initial sequence of images.

[0049] Generally speaking, the applications of the eye tracking method are not limited to these examples. Those skilled in the art will understand more generally that the eye tracking method according to the invention constitutes only one application among others of the method according to the invention. The latter can be used in a number of practical applications, such as monitoring the eyes of a driver of a vehicle, facial recognition, for example in video-conferencing software filters, the detection of tumors in images, for example RX, via CT and / or CBCT scan.

[0050] The present invention also provides: a (computer) data processing system comprising means configured to implement the motion detection method according to any one of the embodiments of the invention; a computer program comprising instructions which, when the computer program is executed by a computer, cause the latter to implement the method according to any one of the embodiments of the invention; a (computer) medium readable by a computer, and on which the aforementioned computer program is recorded.

[0051] The data processing system includes, for example, at least one hardware or computer hardware from among: a computer with a processor for executing algorithms, preferably of PC-type architecture or embedded-type architecture (ARM); an ASIC integrated circuit specific to an application and / or an algorithm; an integrated circuit (logic) of the FPGA type, which is typically reconfigurable or reprogrammable after manufacture.

[0052] The data processing system more preferably consists of an embedded system. This embodiment is advantageously made possible because steps (i) to (iii) of the method according to the invention require low computing power, but also and above all because they make it possible to significantly reduce the number of calculations and the computing power necessary for the execution of the subsequent steps of the method, in particular when these include processing of the images extracted in step (iii). Said embedded system is typically provided with an integrated circuit of the FPGA type.

[0053] The computer-readable computer medium preferably consists of at least one computer medium (or a set of such media) capable of storing digital information. It comprises, for example, at least one of: a digital memory, a server, a USB key, a computer. It may be in a data cloud.

[0054] The embodiments and advantages of the method according to the invention are transposed mutatis mutandis to the computer data processing system, to the computer program and to the computer-readable computer medium according to the invention.

[0055] The invention also proposes a device for detecting a movement of a patient's eye comprising a vision chamber for the patient comprising, at a first end, an opening rim capable of being in contact with a part of the patient's head, and at a second end, a screen for displaying visual stimuli. The screen is integrated into a computer unit comprising: the computer system according to the invention, typically an embedded system equipped with an integrated circuit of the FPGA type; at least one camera for acquiring images of the eye directed towards the first end and electronically coupled to the computer system to transmit to it a sequence of images of the eye. The patient typically places his head on the opening rim so as to direct his gaze into the viewing chamber, and in particular towards the second end within which the screen is arranged. The opening rim is for example in contact with the forehead and nose of the patient. The viewing chamber thus provides a semi-enclosed environment isolated from external disturbances and suitable for detecting and / or tracking the movement of one or more eyes of the patient subjected to visual stimuli, for example for the purposes of a neurological assessment. Preferably, the device comprises at least one first such camera directed towards a first eye of the patient and at least one second such camera directed towards a second eye of the patient to detect and / or track the movement of both eyes of the patient.

[0056] As the device comprises said embedded system, it is made possible to implement the method and / or process according to the invention in real time on the stream of video images acquired by the camera. This embedded technology makes the device according to the invention efficient in detecting and / or tracking a movement of the eye or eyes of a patient.

[0057] The viewing chamber typically comprises a side wall extending about an axis, between the first and second ends. Preferably, a distance between the first and second ends is adjustable to a fixed length of this side wall, measured along the axis, among different possible lengths, these being obtained by successively removable attachments of one or more removable side wall portions to a fixed side wall portion extending from the first end.

[0058] This embodiment of the device makes it possible to modify the distance between the first and second ends, therefore between the patient's eye and the screen, when the device is in operation, for example, during a neurological assessment of the patient, in order to be able to detect and / or follow the movement of the eye according to several conditions of distance from the visual stimuli. Indeed, to modify this distance, it is sufficient to fix or remove one or more of the removable portions, which is particularly easy. Said removable fixings function in particular like a stack of LEGO ® bricks. The fact that each length is fixed allows effective calibration of the device before its use but above all good use of the device in a standardized environment of known dimensions.

[0059] According to one embodiment of the device of the invention, the computer unit further comprises a central clock configured to control: a display of visual stimuli, by means of the screen, by refreshing pixels of the screen at a first frequency; an acquisition of images of the eye, by means of the camera, at a second frequency; the first and second frequencies being equal to or multiple of each other by a non-zero natural number.

[0060] This embodiment implemented at the computer unit level advantageously makes it possible to detect and / or track the patient's eye movement more precisely. Indeed, it is possible to determine precisely to which refreshed pixel(s) of the screen corresponds an image captured by the camera. There is therefore a temporal certainty in the correspondence of the data provided by the screen and the camera because the clock with which the computer unit is provided controls, by itself, the value of the first and second frequencies. In other words, the central clock controls the triggering of the refreshing of the pixels of the screen and the capture of the images by the camera in a synchronized manner. The choice of the frequency values ​​as being integer multiples of one another makes it possible to maintain the temporal relationship between the acquisition of the images by the camera and the presentation of the visual stimuli by the screen. Brief description of the figures

[0061] Other characteristics and advantages of the present invention will appear on reading the detailed description which follows, for the understanding of which reference will be made to the appended figures among which: there figure 1 illustrates an example of a graphical representation of a determination of a similarity measure according to step (i) of the method of the invention; the figure 2 illustrates an example of a graphical representation of a determination of a function derived from the similarity measure of the figure 1 according to step (ii) of the method of the invention; the Figures 3A and 3B illustrate two three-dimensional schematic views of a viewing chamber of an embodiment of the device according to the invention; the figure 4 illustrates a schematic view of a computer unit of a vision chamber of an embodiment of the device according to the invention.

[0062] The drawings of the figures are not to scale. Like elements are generally denoted by like references in the figures. For the purposes of this document, identical or similar elements may bear the same references. Furthermore, the presence of reference numbers or letters in the drawings cannot be considered as limiting, including when these numbers or letters are indicated in the claims. Description of embodiments of the invention

[0063] This section presents a description of preferred embodiments of the present invention. The latter is described with particular embodiments and references to figures but the invention is not limited by them. In particular, the drawings or figures described below are only schematic and are not limiting.

[0064] THE Figures 1 and 2illustrate a concrete example of execution of steps (i) to (iii) of the method of the invention on a sequence of video images of an eye captured by a camera. The images are captured by the camera at 3.75 ms intervals. They are each associated with a discrete probability distribution defined at a pixel of the image as the ratio between the gray level at this pixel evaluated in a discrete manner on a scale of 0 to 255 and the sum of the gray levels at each pixel of the image evaluated in the same way.

[0065] The sequence of image pairs considered in performing step (i) consists of each pair formed by two consecutive images in the image sequence. The similarity measure of each image pair that is considered is that defined by the Bhattacharyya coefficient (BC) of the two discrete probability distributions associated with the two images in the image pair.

[0066] There figure 1represents a graph of the evolution of this similarity measure 2. The reference 2 is used both to designate the similarity measure and its graph. Axis 11 indicates an order number of the images in the sequence. More precisely, each number corresponds to the number of the image having the highest rank in a pair of images, and since the sequence of pairs of images is formed by a succession of two consecutive images, a numbering of the pairs of images returns to that of the images in the initial sequence. Axis 12 is the one on which the value of the Bhattacharyya coefficient of each pair of images is read.

[0067] In order to eliminate the static component of the graph of the figure 1and to eliminate the dependence of the Bhattacharyya coefficient on possible image noise, a derivative function 3 of the similarity measure 2 is used on the sequence of image pairs, according to step (ii) of the method, combined with a low-pass filter. More precisely and preferably, a filtered derivative is applied to the similarity measure 2.

[0068] There figure 2represents a graph of the evolution of such a derivative function 3. The reference 3 is used here both to designate the derivative function and its graph. The filtered derivative considered for this example is more precisely a bilinear discretization of the Laplace transform with transfer function H(s) = s (sτ+1) -1< where τ is a time constant corresponding to five times the time interval between the capture of two images, i.e. 18.75 ms. This example of a filtered derivative is introduced in the description of the invention, in particular in explicit and discretized form. The value of the derivative function 3 in each pair of images can be read on axis 13.

[0069] As the person skilled in the art will easily see in light of the Figures 1 and 2 , applying the derivative function 3 to the similarity measure 2 makes it possible to highlight the pairs of consecutive images in which a significant change in terms of pixels has taken place.

[0070] The points of the graph of the derivative function 3 are filtered to keep only those 31 having a value read on the axis 13 above a strictly positive threshold 4 as illustrated in figure 2. The threshold of 0.00175 introduced in the disclosure of the invention is used for this purpose. Step (iii) of the method is then executed and corresponds to the extraction of the image 51 of the highest rank among the images of each pair of images associated with a point 31 of the graph. This makes it possible on the one hand to detect a clear movement of the eye, approximately between images 545 and 564, without analyzing the images one by one. It is thus possible to analyze this movement more precisely by subsequently processing these images 51, for example via the segmentation step and / or steps (c) and (d) of the eye tracking method described in the disclosure of the invention, in particular to detect eye saccades, which makes such an analysis more efficient and rapid and without requiring a large computing power.The method can further be reinforced by the automatic selection, at a certain frequency, for example 30 Hz, of images 52 from the sequence of images which will be considered, with the images 51, during possible subsequent processing.

[0071] The method according to the invention can be implemented by a "computer" in the generic sense, for example by an embedded system equipped with an integrated circuit of the FPGA type which has the advantage of being able to be incorporated into a portable element of limited power since the method is particularly efficient. A preferred example of a portable element is illustrated in figure 4. This is a computer unit 7 comprising, in addition to the aforementioned embedded system, a screen 71, two cameras 72 and two pairs of infrared emitters 73 arranged symmetrically on either side of the screen 71. The computer unit 7 is configured so that the screen 71 displays visual stimuli, for example, moving pixels, while the cameras 72 capture images of an eye or eyes of a patient. The embedded system can then execute the method according to the invention on the basis of the sequence of images captured by one or both cameras to detect movements of the eye or eyes, for example to detect ocular saccades, and optionally process the images extracted from the sequence of images via step (iii), for example, by segmentation as described in detail in the disclosure of the invention.

[0072] The computer unit 7 is intended to be removably attached to a second end 62 of a vision chamber 6 illustrated in Figures 3A and 3B. The viewing chamber 6 comprises a side wall 8 of approximately conically deformed rectangular parallelepiped shape and extending around an axis Z. A first end 61 of the viewing chamber 6 is opposite the second end 62, the axis Z being directed from the first 61 to the second 62 end. An opening rim 63 is provided at the first end 61 to be in contact with a portion of the patient's head. When the viewing chamber 6 is in use, the patient's forehead rests against an upper portion of the opening rim 63, and a portion of the patient's face between his eyes and his mouth, at the level of his nose, rests against a lower portion of the opening rim 63 opposite the upper portion, a recess being provided in the lower portion for the patient's nose.The patient's eyes are oriented towards the inside of the viewing chamber 6, along the Z axis, towards the screen 71 when the computer unit 7 is fixed at the second end 62. For the readability of the . Figures 3A and 3B , the viewing chamber 6 is shown without the computer unit 7.

[0073] The viewing chamber 6 provides a calibrated environment suitable for the cameras to capture images of the patient's eye or eyes. The infrared emitters 73 are particularly useful for illuminating the patient's eyes for the purpose of this image capture since the environment of the viewing chamber 6 is semi-enclosed and dark. The captured images are thus preferably infrared images.

[0074] The side wall 8 of the viewing chamber 6 consists of a fixed portion 81 and a removable portion 82 which is removably fixed to the fixed portion. Removable fixing elements 64 are provided for this purpose at the end level of the fixed portion 81 opposite the first end 61 and at the ends of the removable portion 82. These fixing elements 64 consist of protrusions and recesses (or more generally reliefs) which cooperate together to establish said removable fixing between the fixed portion 81 and the removable portion 82, in a manner similar to the fitting together of two LEGO ® pieces. Advantageously, similar fixing elements 74 are provided on the computer unit 7, for example around the screen 71, to cooperate with the fixing elements 64 of the removable portion 82 at the second end, and thus removably fix the computer unit 7 to the viewing chamber 6. Such a system is very simple and practical.

[0075] The removable portion 82 makes the assembly dynamic and makes it possible to modify the distance separating the first end 61, therefore the patient's eyes, and the screen 71. In particular, it makes it possible to offer two lengths of side wall 8 measured along the Z axis, fixed and calibrated. A first length is obtained by removing the removable portion 82 and removably fixing the computer unit 7 on the fixed portion 81 via the fixing elements 74 of the computer unit and those 64 at the end level of the fixed portion 81. The second length is obtained with the presence of the removable portion 82, as shown in Figures 3A and 3B. Of course, other similar removable portions, but optionally not all having the same length measured along the Z axis, can be provided in order to offer more than two possible side wall lengths 8, that is to say several possible distances between the patient's eyes and the screen. These removable portions 82 can be successively fixed to each other in a removable manner on and from the fixed portion 81, thanks to the fixing elements 64, thus generalizing the assembly of the Figures 3A and 3B The computer unit 7 is then placed at the second end 62, the position of which depends on the number of removable portions 82.

[0076] Those skilled in the art will understand that these characteristics on the viewing chamber 6 and its attachment to the computer unit 7 can also be considered independently of the method according to the invention and of the fact that the computer unit 7 comprises means for implementing this method.

[0077] In summary, the invention relates to a computer-implemented method for detecting a movement of a body from a sequence of images of this body, by isolating (or extracting) the images 51 belonging to pairs of images in which a derived function 3 of a similarity measure 2 of these pairs of images is greater than a predetermined threshold 4. In this way, the movement of the body is detected on these images 51, which can then be processed more precisely for the purpose of studying the movement, without requiring processing all the images of the sequence of images.

[0078] The present invention has been described above in relation to specific embodiments, which are of purely illustrative value and should not be considered as limiting. It will be readily apparent to those skilled in the art that the invention is not limited to the examples illustrated or described above, and that its scope is more broadly defined by the claims introduced below.

Claims

1. A method for detecting a movement of a body from a sequence of images of this body, the method being computer-implemented and comprising the following steps: (i) determining a similarity measure (2) of each image pair of a sequence of image pairs, the sequence of image pairs being induced by the sequence of images and including each image thereof; (ii) determining a derivative function (3) of the similarity measure (2) on the sequence of image pairs, the derivative function (3) corresponding to a filtered derivative obtained by combination with a low-pass filter; (iii) extracting at least one image (51) from each image pair in which the derivative function (3) is greater than a positive threshold (4); (iv) detecting the movement of the body on the basis of the images extracted in the step (iii).

2. The method according to claim 1, wherein the similarity measure (2) determined in the step (i) corresponds to a Bhattacharyya coefficient of discrete probability distributions associated with the images of the image pair and defined from pixels of these images.

3. The method according to claim 2, wherein the discrete probability distribution associated with each image is defined at a pixel of the image as a quotient of a grey level at that pixel by a sum of grey levels at each pixel of the image.

4. The method according to any one of claims 1 to 3, wherein the filtered derivative corresponds to a bilinear discretisation of a transfer function Laplace transform H(s) = s (sτ+1)-1 where τ is a constant.

5. The method according to any one of claims 1 to 4, wherein the threshold (4) is defined by a percentile of a distribution of data obtained by applying the derivative function (3) to the sequence of image pairs.

6. The method according to claim 5, wherein the percentile is chosen between the eightieth and the ninety-ninth percentile.

7. The method according to any one of claims 1 to 6, wherein the step (iii) also comprises periodically extracting an image (52) from the sequence of images.

8. The method according to any one of claims 1 to 7, wherein said at least one image (51) extracted in the step (iii) comprises the image of highest rank in the sequence of images among the two images of the image pair.

9. The method according to any one of claims 1 to 8, wherein the sequence of image pairs comprises each pair of two consecutive images in the sequence of images.

10. The method according to any one of claims 1 to 8, wherein the sequence of image pairs comprises each pair formed of the first image of the sequence of images and a further image of the sequence of images.

11. The method according to any one of claims 1 to 10, comprising a segmentation step of the images extracted in the step (iii).

12. The method according to claim 11, wherein the segmentation of each image (51, 52) extracted in the step (iii) is defined by a segmentation function of the image, and wherein the segmentation step is implemented via the following sub-steps: (s1) determining an estimate of the segmentation function of each image (51 52) extracted in the step (iii); (s2) recursively determining the segmentation function of each image (51, 52) extracted in the step (iii) on the basis of its estimate.

13. The method according to claim 12, wherein the estimate of a segmentation function of a second image is determined in the substep (s1) by a composition of a vector field with a segmentation function of a first image, and wherein - the first image precedes the second image in the sequence of images, - the segmentation function of the first image is determined beforehand in the sub-step (s2), and - the vector field corresponds to a displacement of pixels between the second and first images.

14. An eye-tracking method comprising the following steps: (a) providing a sequence of images of an eye of a patient; (b) detecting a movement of the eye from the sequence of images; wherein the step (b) is carried out by means of the method according to any one of claims 1 to 13.

15. The method according to claim 14, wherein the step (b) is implemented by means of the method according to any one of claims 11 to 13, the segmentation step being implemented at least in the vicinity of a representation of the iris of the eye on each of the images (51, 52) extracted in the step (iii), the method further comprising the following steps: (c) determining a position of a limbus of the eye on the basis of the segmentation of the images (51, 52) extracted in the step (iii); (d) determining a position of the eye on the basis of the position of the limbus of the eye determined in the step (c).

16. Use of the method according to claim 14 or 15, to detect a saccade of the eye.

17. A data processing computer system comprising means configured to implement the method according to any one of claims 1 to 13.

18. The computer system according to claim 17, consisting of an embedded system equipped with an FPGA-type integrated circuit.

19. A device for detecting a movement of an eye of a patient comprising a viewing chamber (6) for the patient comprising, at a first end (61), an opening periphery (63) capable of being in contact with a head part of the patient, and at a second end (62), a screen (71) for displaying visual stimuli; the screen (71) being integrated into a computer unit (7) comprising: - the computer system according to claim 18; - at least one camera (72) for acquiring images of the eye directed towards the first end (61) and electronically coupled to the computer system in order to transmit to the latter a sequence of images of the eye.

20. The device according to claim 19, wherein the viewing chamber (6) comprises a side wall (8) extending about an axis (Z), between the first (61) and the second (62) ends, a distance therebetween being adjustable to a fixed length of the side wall (8) measured along the axis (Z) from among different possible lengths, the different possible lengths being obtained by successive removable attachments of one or more removable portions (82) of the side wall (8) to a fixed portion (81) of the side wall (8) extending from the first end (61).

21. The device according to claim 19 or 20, wherein the computer unit (7) further comprises a central clock configured to control: - a display of the visual stimuli, by means of the screen (71), by refreshing pixels of the screen (71) at a first frequency; - an acquisition of images of the eye, by means of the camera (72), at a second frequency; the first and second frequencies being equal to or multiples of each other by a non-zero natural number.

22. A computer program comprising instructions which, when the computer programme is run by a computer, lead the latter to implement the method according to any one of claims 1 to 13.

23. A computer-readable medium on which a computer program according to claim 22 is recorded.